Decentralizing Science with Knowledge Graphs
summary
Platform moves from data governance to an AI-native distributed system that trains domain-specific LLMs on proprietary studies while preserving privacy with fully homomorphic encryption. Experiments run with Harvard MIT and Oxford inform a live prototype that converts papers into node/edge knowledge maps — dynamic knowledge graphs updated continuously by paid contributors — and pays royalties in tokens when data is reused. First product is Gin Tonic marketplace to repurpose idle GPUs for AI compute, issuing GIN tokens as rewards.
AI-written summary · about
chapters (9)
- 0:00Introduction to vision
- 0:49Data governance challenges
- 2:11Building the platform
- 3:58AI-native distributed platform
- 6:05Why knowledge graphs
- 9:36Web3 and knowledge graphs
- 10:40Applications in longevity
- 12:03Team and mission
- 12:48Decentralized science accelerator
related talks
19mWisdom-Driven Knowledge Augmented Generation at Scale
42mMaking OSS AI Real with Knowledge Graphs
41mEnhancing LLMs with Graph Technology
28mThe Present and Future of AI-Enabled Knowledge Graph Generation
25mEnhance LLM Applications' Trustworthiness and Explainability
30mBuilding Knowledge Graphs With LLMs
14mKnowledge Graphs as the Backbone of AI-First Software Delivery
52mGive Your LLMs a Left Brain